This reverts commit 63d786cf22.
These functions will be used for third party OTA zips, so keep them.
Change-Id: I24b67ba4c86f8f86d0a41429a395fece1a383efd
It seems that old Scrypt-config.mk is unaware of arm64 architecture.
In fact, crypto_scrypt-neon.c can be compiled into arm64 as well. This
CL adds an arch section for arm64.
Bug: 65425184
Test: adb shell /data/nativetest64/scrypt_test/scrypt_test
Change-Id: Ib451de642c7cc4548bfdc0879781981654b21b8f
Merged-In: If40a30378b8038324aad44071107130d7722e28d
Add debug capability.
Fix backup of images after tarred partitions with too much padding
Add more md5 logging.
Skip digest check if selected.
Change ADB Backup version to 3 for new fixes to image padding.
Change-Id: I0f76c0733c523717e4797d1a14c3ae47d046fc8c
Enable a property when mounting partitions required for decryption, and disable when they are unmounted.
This helps init scripts to react to such an event appropriately, for example by starting services critical for decryption that reside on said partitions.
Change-Id: Ib6fa1e5c903d9c800fb145d582ced6cc7e0ab5b9
- Add util.UpdaterStates - the state of SystemUpdaterSample.
It's different from status of UpdateEngine;
when UpdateEngine#cancel is used to suspend the update,
UpdateEngine sets status to IDLE, which cannot be used
to track the suspended state.
- UI: Change 'Update status' to 'Engine status'.
- UI: Change 'Update completion' to 'Engine error code'.
- UI: Add 'Updater state'.
Test: manually on the device
Test: using JUnit4
Change-Id: I9c58b5ed0eae3be7ab8b217fc01a621e8fb2f4bf
Signed-off-by: Zhomart Mukhamejanov <zhomart@google.com>
* Allow custom definition of strace path with TW_ flags.
* `TW_CRYPTO_SYSTEM_VOLD_DEBUG := true`
will use default path: /sbin/strace.
* `TW_CRYPTO_SYSTEM_VOLD_DEBUG := /system/xbin/strace`
will use the provided `/system/xbin/strace` path.
Change-Id: I5e12a10176d17a4f26487de0976a776d48c4142e
Signed-off-by: nailyk-fr <nailyk_git@nailyk.fr>
* Separate stdout and stderr buffers:
vdc's return codes get sent to stdout, but the possible presence of
other error messages in the output buffer will cause a valid return
from vdc not to be parsed properly, and subsequent decryption to fail
due to "misunderstood" return code.
eg on the U11+ (htc_ocm) libc will generate an error to stderr due to
a missing property area resulting a proper connection to vold being
incorrectly parsed, and breaking decryption.
* Improve logging.
Change-Id: I57987ebe4ee6754a78e79ca177506098f8301f8f
For devices with the Qualcomm RTC fix whose clocks need further tuning
such as the LG V20 whose ats files produces the correct time except
46 years in the future.
Takes adjustment in seconds and adjusts after reading the the ats file.
Boardconfig.mk
TW_CLOCK_OFFSET := -1451606400
Change-Id: I0b2d35dd0026ed18a7bbebf683517f7c0eaa53ff
For example on my device, the LG V20 (US996), the default time
comes back as 2016 which doesn't trigger the fixup code.
The actual time I picked was pretty arbitrary, but close to
the current date, since we're checking if said date/time is reasonably
accurate without the fixup code.
Change-Id: I01db64777d7d0c5ce0710445cf88858d89656ac0
Decrypting userdata is not needed when installing zips if we have
a map of blocks to read, so skip the decryption. This allows to
install zips with no user interaction and, more importantly, it
allows to install zips on devices where TWRP can't properly decrypt
userdata.
Remove also the unneeded Mount_All_Storage() call, Install_Command()
already mounts the needed partition.
Change-Id: I91c69f84ecc2fe41c9567c40649e501af1cfb64a
if we try and flip pages too quickly, the kernel returns -EBUSY.
Use drmHandleEvent to wait for any previous flip to complete
before attempting a new one.
Change-Id: If3def889c0c89d7cfb8b7f13c4dae6d4b12760f9
draw performance is significantly slower when drawing directly to the
mmaped buffer. Instead, draw to a userspace buffer and memcpy()
to the mmaped buffer on page flip.
Change-Id: Ibf2cd4be4f5437b445e68c572ff76ad72d506b15
Otherwise tests may interfere with each other by using the same / default
location.
Test: Run recovery_component_test on marlin.
Change-Id: I6b0455489f2fdce819009964dd92bfd9bfeb06ef
The imgpatch will fail on empty deflates because the bspatch won't call
the sink function if the target length is zero.
Instead of compressing an empty string, it's cleaner to not generate such
empty deflate chunks in the patch. Therefore, we can just convert the
chunk type to raw if the target length is smaller than the patch data.
Also adjust some unit tests and add the testdata gzipped_source &
gzipped_target. These two files are ~1K each and are generated by
gzipping two slightly different regular files.
Bug: 79265132
Test: unit tests pass, imgpatch applys successfully on the given src/tgt
Change-Id: I6bfff3251918137f6762a6f9e9551642371a1124
Move the common codes into RunBlockImageUpdate(). Also clean up the
partition updated marker after running each test.
Test: Run recovery_component_test on marlin.
Change-Id: Id4302e4da4c664231b737a1e83d2e164ef58ed97
Libraries that are direct or indirect dependencies of modules installed
to recovery partition (e.g. adbd) are marked as recovery_available:
true. This allows a recovery variant of the lib is created when it is
depended by other recovery or recovery_available modules.
Bug: 79146551
Test: m -j
Change-Id: I9ca07ef53fbd89e7c86c0ba269fa52bb48d6deb4
- Add UpdateManager - responsible for the update logic.
Now ui.MainActivity is responsible for only UI.
- Create sample test for UpdateManager
- Remove MainActivityTest - now MainActivity is really simple.
- Add separate callback for progress update.
- MainActivity: UpdateEngine#bind/unbind on pause/resume.
Test: manually on the device
Test: using JUnit4
Change-Id: I1dba7c4ec74b1afb520be762413cfc261ccfbc08
Signed-off-by: Zhomart Mukhamejanov <zhomart@google.com>
Tao Bao
Zhomart Mukhamejanov
Treehugger Robot
Tom Marshall
Logan Chien
bigbiff bigbiff
oshmoun
nailyk-fr
nkk71
Phoenix591
Gabriele M
Simon Shields
Jiyong Park
Tianjie Xu